LAUSR

Partially massless (PM) fields of spin-2 have been subject to renewed attention in recent years after the important advances made in the understanding of massive gravity [1,2]. In four-dimensional de Sitter (dS4) spacetime with (positive) cosmological constant Λ , a PM graviton has a mass given bymPM = 2Λ/3 and a corresponding gauge invariance that removes the degree of freedom associated with the spin-zero mode of the particle. A consistent theory of PM gravity would be very attractive given its relevance in the context of cosmology and the fact that a PM spin-2 field is not subject to the same strict experimental constraints as a generic massive graviton, e.g. the bounds on fifth-force experiments or on dispersion of gravitational waves. In spite of their potentially interesting phenomenology, complete and physically realistic models involving PM fields are currently lacking. A crucial hurdle one encounters when attempting to construct such a PM theory beyond linear level is the requirement of gauge invariance of the interactions. Indeed, this condition is restrictive enough to rule out theories of a single self-interacting PM spin-2 particle, see e.g. [3–5]. Such no-go results beg the question of whether a fundamental obstruction exists for the construction of non-trivial PM models. A first step in order to address this question is to precisely understand what are the assumptions that lead to the existing negative results. In the recent work [6] we have shown that, by relaxing the requirement of classical unitarity, one can in fact construct a complete theory for a multiplet of PM gravitons around (A)dS4 space, as we review in the following. This is an interesting outcome as it